With the rapid urbanization, the increasing crime is the main challenge of urban governance. As a new governance model, smart city aims to alleviate various urban diseases. However,can constructing a smart city reduce crime? There is no evidence for this. Based on the quasi natural experiment of smart city pilot policy launched in 2012 in China, this paper collects the paneldata of 292 prefecture level cities from 2005 to 2020, uses the staggered difference-in-difference modelto evaluate the effect of smart city on crime, and makes an in-depth discussion on its mechanism.The study found that: (1) Building a smart city can significantly reduce the number of crimes; (2)Smart city mainly exerts its influence on reducing crime through three mechanisms: Increase residents?? income, improve the application of technology and optimize the allocation of resources; (3) Heterogeneity analysis shows that cities of different sizes and levels of developmenthave different effects on crime reduction through the smart cities. This paper provides important policy enlightenment for improve the urban public security.

伴随快速城镇化不断增加的犯罪是当下城市治理面临的主要挑战。智慧城市作为一种新型治理模式,其目的是纾解快速城镇化引致的各类“城市病”。伴随着实践层面的不断推进,建设智慧城市能否减少犯罪发生? 这一兼具学理性与实践性问题被提出。遗憾的是,对此问题,学界目前尚未展开深入、细致地讨论,尤其缺少基于面板数据的针对性研究。鉴于此,本研究将2012年我国启动智慧城市试点政策视为准自然实验,收集292个地级市2005—2020年的面板数据,使用渐进双重差分模型评估建设智慧城市对犯罪发生的抑制效应,并就其中的作用机制进行深入剖析。研究发现:(1)建设智慧城市能够显著减少当地犯罪发生;(2)机制分析表明,智慧城市主要通过增加居民收入、提高技术应用与优化资源配置三个机制抑制犯罪发生;(3)异质性分析发现,不同规模、发展水平的城市,通过建设智慧城市减少犯罪的效果存在差异。本文为改善城市安全提供了重要的政策启示。